Managed care plans and other health care providers face a difficult task in predicting outpatient mental health services use. Existing research offers some guidance, but our knowledge of which factors influence use is confounded by methodological problems and sampling constraints. Consequently, available findings are insufficient for developing accurate predictions, which managed care plans need in order to formulate fiscally responsible service delivery contracts. This article reviews the primary data sources and research on ambulatory mental health services. On the basis of this review, the probability and intensity of outpatient visits are estimated. The primary predictors of use are also examined because they may help managed care plans forecast use by a given population or group of enrollees. Gender, age, race, education, health status, and insurance coverage are several variables surfacing as statistically significant predictors of use. The implications for planning capitated mental health services are discussed.